Amidst allegations of OpenAI mimicking Scarlett Johansson’s voice for its ChatGPT’s “Sky” avatar, internal documents obtained by The Washington Post reveal otherwise. Contrary to Johansson’s apprehensions, the AI giant did not intentionally replicate the actor’s voice for its synthetic avatar.
The controversy arose when Scarlett Johansson expressed shock over a synthetic voice released by OpenAI, claiming it bore an unsettling resemblance to her own. However, OpenAI CEO Sam Altman’s public apology on May 21, 2024, was accompanied by revelations from internal documents suggesting a different narrative.
These documents shed light on the meticulous process behind selecting the voice for ChatGPT’s Sky avatar. Contrary to speculation, OpenAI had engaged a different actress to voice Sky months before any discussion with Scarlett Johansson took place.
Insights from interviews with individuals involved in Sky’s voice creation corroborated OpenAI’s stance. The company had not actively sought a voice resembling Johansson’s during the casting process. Instead, criteria focused on attributes like warmth, engagement, and charisma, with no explicit mention of Johansson or her work in the movie “Her,” where she voiced an AI character.
These revelations provide clarity amidst the controversy, dispelling claims of deliberate mimicry and reaffirming OpenAI’s commitment to ethical AI development.
